This project aims at improving the current knowledge and modeling of snow at microscale in order to guide the development of new generation snowpack models, which are necessary for several applications such as climate modeling, avalanche risk forecasting, etc. All the critical ingredients involved in snowpack modeling, namely metamorphism, transfer, mechanical and optical properties of snow, will be revisited and developed further within original multiscale approaches.
The project aims to enable the development of new generation snow models by addressing 6 key points of snow modeling from its 3D microstructure:<br />-Task 1: isothermal metamorphism<br />The objective here is to develop a digital micro-scale model which describes the evolution of snow in 3D under isothermal conditions and which fully accounts for the main specificities of such a metamorphism, in particular the rearrangement of particles, caused by the sublimation of the bridges between grains, due to Kelvin's law.<br />-Task 2: gradient metamorphism<br />The aim is to develop a digital model of micro-scale gradient metamorphosis, taking into account in particular the effects of faceting. It will be evaluated and improved using time-lapse images acquired by tomographic methods.<br />-Task 3: wet snow metamorphism<br />The main objective of this task is to develop an experimental method to obtain successive 3D images of wet snow samples clearly showing the 3 phases: ice, pore and liquid water. A complementary objective is to adapt the phase field model developed in T1 to model the metamorphism of snow saturated in liquid water.<br />-Task 4: transfer properties<br />The objective of this task is to better quantify and model the influence of snow metamorphism on the effective transfer properties of snow (thermal conductivity, water vapor diffusion, permeability, source term due to sublimation-deposition processes). In particular, the homogenization work of Calonne et al (2015) previously carried out will be applied to one or more controlled temperature gradient experiments in an instrumented chamber.<br />-Task 5: mechanical properties<br />The work will mainly focus on the development of a dual-scale MPMxDEM numerical simulation tool that will make it possible to perform macro-scale mechanical simulations from a micro-scale DEM description, and thus fully take into account the knowledge available on the snow microstructure. In-situ mechanical tests will also be carried out to assess and validate digital developments.<br />-Task 6: Mechanical properties<br />This task proposes the development, validation and application of an efficient and easily applicable Ray-Tracing model on 3D snow images. It will be validated using several experimental approaches.
As suggested with the Greek term «mimesis«, the project aims at imitating the microscale behavior of snow using numerical simulation to better understand and describe the underlying snow physics. It is based on the study of snow metamorphism and its properties using systematic comparisons between 3D numerical modeling and microscale experimentation. For this purpose, we will take advantage of experimental results to validate and refine the models we have (e.g. Bretin et al, 2015; Flin et al, 2018) or will develop. Concerning experimentation, we will benefit from our room temperature cold cell (Calonne et al, 2015), which allows monitoring dry snow metamorphism directly under X-ray tomography or DCT (voxel size around 8 um). In addition, the OSUG, INSU & CNRM have recently acquired a coldroom tomograph (TomoCold), which will be made available to the cryospheric community. This new instrument has recently been validated down to -30°C and will be especially suited to tasks T4 to T6. Most of the results, valid at the grain or layer scale, will then be used to infer the snow behavior at the snowpack scale using multiscale approaches (see e.g. T4 and T5).
Direct scientific outcomes of MiMESis-3D will consist in a better knowledge of snow, its metamorphism, and the evolution of its properties over time. More specifically, T1 and T2 will allow the determinant physical mechanisms involved in snow metamorphism to be clearly identified, leading to:
-1) accurate numerical simulation tools to investigate snow metamorphism at microscale for a continuum of experimental conditions.
-2) crucial results for the homogenization (multi-scale) approaches at the snowpack scale by confirming or infirming the equations and physical quantities mainly involved.
Task T5 will yield experimentally and numerically validated 3D constitutive laws (or reduced interface laws) that could be integrated in macro-scale models for avalanche risk forecasting. Task T6 will provide a numerical tool to investigate the relationships between the microstructure and the optical properties of snow, which are important, e.g., for an accurate snow characterization by optical methods (SSA estimations) and remote sensing. If successful, task T3 will provide powerful experimental tools and methodological approaches to study wet snow metamorphism.
The combination of all the tasks will give way to systematic computations of the evolution of snow properties under well-defined experimental conditions, thus helping in better understanding the microstructure-properties relationships.
One of the most direct application of this project is a full multiscale modeling of the snowpack physics, which will lead to a more accurate estimation of snowpack properties, an improved understanding of climate feedbacks (CES01) and a better knowledge of avalanche triggering. Among several models, Crocus, the open-source snowpack model of Météo-France, will significantly benefit from such a research. An improved model can also provide interesting results for remote sensing and data assimilation of snow cover. From a longer and broader perspective, snow is a mixture of air and water, which are probably the most essential elements of the climate system. With issues such as global warming in which polar regions play a key role, fostering fundamental scientific knowledge on such crucial components at low temperature seems a useful step toward larger scale climate-related research.
Dumont, M., Flin, F., Malinka, A., Brissaud, O., Hagenmuller, P., Lapalus, P., Lesaffre, B., Dufour, A., Calonne, N., Rolland du Roscoat, S., and Ando, E.: Experimental and model-based investigation of the links between snow bidirectional reflectance and snow microstructure, The Cryosphere, 15, 3921–3948, doi.org/10.5194/tc-15-3921-2021, 2021.
Granger, R., Flin, F., Ludwig, W., Hammad, I., and Geindreau, C.: Orientation selective grain sublimation–deposition in snow under temperature gradient metamorphism observed with diffraction contrast tomography, The Cryosphere, 15, 4381–4398, doi.org/10.5194/tc-15-4381-2021, 2021.
Bretin, E., Denis, R., Masnou, S., Sengers, A., Terii, G., A Cahn-Hilliard multiphase system with mobilities for the simulation of wetting arxiv.org/abs/2105.09627, preprint.
Bretin, E., Masnou, S., Sengers, A., Terii, G., Approximation of surface diffusion flow: a second order variational Cahn-Hilliard model with degenerate mobilities arxiv.org/abs/2007.03793v1, accepted M3AS.
Communications in conferences :
Flin, F., Granger, R., and Geindreau C. Towards the simulation of temperature gradient snow metamorphism: a phase field approach, Surface and Interface Dynamics II, Tokyo, Japan, 21 22 October, 2020, invited talk (videoconferencing).
Bouvet, L., Calonne, N., Flin, F., and Geindreau, C.: Modeling snow isothermal metamorphism at the pore scale with the phase-field model Snow3D, EGU General Assembly 2021, Vienna, 19-30 April, 2021, vPICO (videoconferencing).
Herny C., Hagenmuller P., Chambon G., Roulle J., Peinke I. : Microstructure-based modelling of snow mechanics : experimental evaluation on the cone penetration test, ALERT Geomaterial Workshop, Aussois, 27-29 September 2021, poster.
Ozenda, O., Chambon, G., and Richefeu, V. Multiscale MPMxDEM method for geophysical flows with a cohesive granular microstructure, ALERT Geomaterial Workshop, Aussois, 27-29 September 2021, oral communication.
Understanding and predicting snowpack evolution is very important for several applications ranging from climate modeling to avalanche risk forecasting, road viability, water resources availability and management. Snow on the ground is a complex porous material that constantly transforms over time. The evolutions incurred during its metamorphism strongly impact the physical and mechanical properties of the snowpack. Current macro-scale snow models take into account the layered and changing structure of snow, but rely on a very limited set of scalar and largely-empirical parameterizations to describe the full complexity of snow microstructure. Preliminary homogenization approaches to bridge micro and macro scales, relying on 3D snow images obtained by X-ray tomography, have recently made great progress at the international level. However, these approaches remain fundamentally “static” and only provide properties based on the actual observed microstructure. A new and great challenge is to add the dynamic component, i.e., to extend the techniques to the metamorphism effects occurring in the snowpack.
The general objective of this project consists in improving the current knowledge and modeling of snow at micro-scale in order to guide the mid-term development of a new generation of snowpack models that better account for metamorphism-induced evolutions of snow microstructure. Through systematic comparisons between 3D numerical modeling and micro-scale experimentations, the project aims at investigating the microscale behavior of snow to better understand and describe the snow physical and mechanical properties at macro-scale. We propose to study the evolution of snow microstructure based on both (1) 3D metamorphism models and (2) images to be acquired by X-ray microtomography in experiments performed under experimentally controlled conditions. In detail, the main technical and scientific objectives are:
-Couple purely thermodynamics and mechanical effects to model the morphological changes of the microstructure during equi-temperature metamorphism. Simulations based on the Discrete Element Method (DEM) will be coupled to existing thermodynamics models in order to properly account for densification effects.
-Develop and evaluate models for temperature Gradient (TG) metamorphism, which is characterized by important faceting effects. The works will first consider simplified configurations of known crystalline orientations. Then, models will be applied to complete snow samples.
- Monitor wet snow metamorphism under X-ray tomography, which is a more exploratory task.
-Develop and apply up-to-date numerical homogenization models to compute snow transfer, mechanical and optical properties on the base of micro-tomography images and outputs of micro-scale metamorphism models. Systematic results obtained from series of microstructures will provide accurate data for the development of macroscopic snowpack models.
After validation, the resulting models will provide simulating tools that are inaccessible today, and awaited by the various scientific communities working on snow from the micro to the macro scales.
To tackle these multidisciplinary issues, the project team gathers scientists originating from diverse but connected scientific communities (applied mathematics, mechanics, material sciences, environmental engineering). Active collaboration between the partners will systematically be promoted, and additional external expertise will also be sought when required.
Monsieur Frédéric Flin (Centre national de recherches météorologiques)
The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.
ETNA EROSION TORRENTIELLE, NEIGE ET AVALANCHES
CNRM Centre national de recherches météorologiques
3SR Sols, Solides, Structures, Risques
IGE Institut des Géosciences de l'Environnement
ICJ Institut Camille Jordan
Help of the ANR 671,693 euros
Beginning and duration of the scientific project: March 2020 - 48 Months